Improved random forest model for coronary heart disease pre-diagnosis and pre-diagnosis system thereof

Abstract

The invention discloses an improved random forest model for coronary heart disease pre-diagnosis and a pre-diagnosis system thereof, which are characterized in that nearly hundreds of indexes used in clinical examination of coronary heart disease patients are screened based on feature selection in combination with a statistical machine learning algorithm, a model with high prediction precision is established, and the model can be used as a basis for whether coronary angiography is carried out or not. The risk that a patient suffers from the coronary heart disease at present is calculated by combining clinical examination indexes, image detection results and electronic medical record information of the patient, and an evaluation basis is provided for subsequent definite diagnosis.

Description

technical field [0001] The invention belongs to the research technology in the field of data training and statistical machine learning data mining, and in particular relates to an improved random forest model used in the prediagnosis of coronary heart disease and a prediagnosis system thereof. [0002] technical background [0003] Heart disease is the disease with the highest fatality rate in the world, especially in my country, and its fatality rate exceeds the sum of all cancers. According to the data given in the 2019 China Cardiovascular Health and Disease Report, overall, the prevalence and mortality of cardiovascular diseases in China are still on the rise. It is estimated that the number of patients with cardiovascular diseases is 330 million, of which 13 million are stroke, 11 million are coronary heart disease, 5 million are cor pulmonale, 8.9 million are heart failure, 2.5 million are rheumatic heart disease, 2 million are congenital heart disease, and 2 million ar...

AI Technical Summary

Problems solved by technology

[0019] First of all, the screening conditions of the 17 biomarkers and the difficulty and cost of obtaining them in clinical testing have not been clearly stated, so the possibility of practical application cannot be guaranteed; The cost and equipment requirements of small or higher precision assays are high

[0020] Secondly, the threshold of the disease probability is limited to 50% in the study, that is, the results obtained are either black or white, but if the diagnosis result is disease, it cannot provide the data characteristics of the degree of depth and contribution

Therefore, the results of the diagnosis give too little information to the patient, and it is impossible to provide a specific basis based on individual differences.

In addition, the threshold limit is not an existing standard in the industry, but an empirical value setting in the data analysis process, so the design of parameters lacks authority and is subjective

Images